Accepted Manuscript

Title: Trust and technology <!–<query id="Q3">Please check

the doc head for correctness.</query>–>transfers

Authors: Marı́a Garcı́a-Vega, Elena Huergo

PII: S0167-2681(17)30207-X
DOI: http://dx.doi.org/doi:10.1016/j.jebo.2017.07.029
Reference: JEBO 4109

To appear in: Journal of Economic Behavior & Organization

Received date: 8-2-2017

Revised date: 21-7-2017
Accepted date: 24-7-2017

Please cite this article as: {http://dx.doi.org/

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Trust and Technology Transfers
By María García-Vega and Elena Huergo*

* García-Vega: School of Economics, University of Nottingham, University Park, Nottingham NG7 2RD, United
Kingdom and GEP (email: maria.garcia-vega@nottingham.ac.uk); Huergo: Department of Economic Analysis.
Universidad Complutense de Madrid 28223, Spain (email: ehuergo@ccee.ucm.es).

Highlights
 Research and development is largely done by multinationals that transfer technology
to their foreign subsidiaries.
 Trust might be an important determinant of the governance of technology transfers
 We empirically investigate how widely held perceptions of the trustworthiness of
the host economy influence technology transfers that subsidiaries receive from their
business group.
 We find that increases in perceived trustworthiness raise technology transfers from
the market and reduce technology transfers from the business group.
 Our results support predictions of transaction cost economics about how technology
transfers are organized.

Abstract

Research and development is largely done by multinationals (MNEs) that transfer

technology to their foreign subsidiaries. Trust might be an important determinant of the
governance of technology transfers because trust can reduce the dependence of the subsidiary
on the headquarters. We empirically investigate how widely held perceptions of the
trustworthiness of the host economy influence international technology transfers that
subsidiaries receive from their business group or from other international providers. We use
firm-level data on R&D imports from foreign subsidiaries operating in Spain for the period 2005
to 2012, and a Eurobarometer measure of trust between citizens of European countries. We find
that subsidiaries that belong to MNEs from countries with higher trust in Spaniards have fewer
technology transfers within the business group and more from international market channels than
subsidiaries from countries with lower trust in Spaniards. Our results support predictions of
transaction cost economics about how technology transfers are organized.

Keywords: Trustworthiness, technology transfers, R&D imports, heterogeneous firms.

* García-Vega: School of Economics, University of Nottingham, University Park, Nottingham NG7 2RD, United
Kingdom and GEP (email: maria.garcia-vega@nottingham.ac.uk); Huergo: Department of Economic Analysis.
Universidad Complutense de Madrid 28223, Spain (email: ehuergo@ccee.ucm.es).

This research has been partially funded by the Spanish Ministry of Economy and Competitiveness (project
ECO2014-52051-R) and by the Autonomous Region of Madrid through project S2015HUM-3417
(INNCOMCON-CM), co-funded by the European Social Fund. We would like to thank Juan Mañez, Simon
Gächter and Sourafel Girma for their very helpful comments along with seminar participants at the University of
Nottingham, University of Granada, 25th EEA Conference (Glasgow), 37th Annual EARIE Conference, 12th
ETSG Conference and CESifo workshop (Venice). We are also grateful to Mary Patricia Tamayo for research
assistance.

1
JEL Classification: D23, L14, L24, O30

1. Introduction

There is ample evidence on the positive effect of trust on economic growth and

development (La Porta et al., 1999; Knack and Keefer, 1997; Zak and Knack, 2001; Dearmon

and Grier, 2009; Guiso et al., 2008). One mechanism for the positive effect of trust on growth

is that trust positively influences international transfers of technology, which in turn can

enhance productivity and growth in the host economy (Eaton and Kortum, 1996; Keller, 2004).

In this paper, we investigate this mechanism. We focus on the international technology transfers

that subsidiaries receive from their business group and from other foreign providers.

Multinationals (MNEs) are key drivers of innovations and are responsible for a large number

of technology transfers (Markusen, 1984; Branstetter et al., 2006; Bertrand, 2009; Criscuolo et

al., 2010). Therefore, understanding the nature of technology transfers from MNEs is of great

economic relevance.

Our specific contribution is to examine whether the trustworthy environment in which

multinationals and their foreign subsidiaries operate impacts the intensity of within-group and

through-market international technology transfers. The hypothesis that trust might reduce

within-firm transactions and increase market channel flows goes back to Williamson (1971).

According to Williamson, “vertical integration would be more complete in a low-trust than a

high-trust culture, ceteris paribus” (p. 122). There is evidence that subsidiaries that operate in

environments perceived as less trustworthy might be more dependent on the headquarters than

subsidiaries operating in highly trusted environments (Bloom et al., 2012). And as a

consequence, they will receive more technology transfers from within the business group and

fewer technological services from international market channels.

2
 To shed light on the role of trust for international technology transfers, we use a measure

of trust from national, representative Eurobarometer surveys, as in Guiso (2009) and Bloom et

al. (2012). In the surveys (details are in Section 3.2), citizens from different countries are asked

how much trust they have in people from various countries (a lot, some, not very much, no trust

at all). This measure of trust is a measure of generalized trust and it reflects the perception of

trustworthiness of the average citizen of a given country by citizens of another country (Guiso

et al., 2009).

How can we interpret such perceptions of generalized trust in light of our research

question? Williamson (1993a) argues that “trust is warranted when the expected gain from

placing oneself at risk to another is positive, but not otherwise. Indeed, the decision to accept

such a risk is taken to imply trust”. Williamson (1993b) makes three conceptual distinctions of

trust by dividing trust into calculative trust (considering potential risks), personal trust (from

repeated interactions) and institutional trust (from the organizational context of the contracts).

Arguably, people’s answers in the Eurobarometer survey reflect all three forms of trust. People

might be calculative in the sense that they weigh up previous personal experiences with citizens

from other countries. However, even if citizens do not have these experiences, they can have a

perception of trustworthiness that might depend on, for example, institutional quality and on

cultural elements. Consistent with this argument, Guiso et al. (2009) find that cultural aspects,

positive expectations of a country’s law enforcement, business morale, and the quality of the

institutions form the beliefs upon which the generalized trust between citizens of countries from

the Eurobarometer is based. Hence, for our purposes, we consider the measure of trust we

deploy as a useful summary statistic for the perceived trustworthiness of a country’s economic

environment. Note that Williamson (1981) considers that trustworthiness can be regarded as

the absence of opportunism. Opportunism is key in transaction costs economics. Indeed,

3
Williamson states that “contracting would nevertheless be feasible if economic agents were

completely trustworthy” (p. 1545).

The Eurobarometer data reveal strong differences in perceived trustworthiness between

nations. For example, Italians view Spaniards as more trustworthy than Dutch people do.

Hence, it is plausible to assume that managers of Italian and Dutch MNEs, who decide on how

to organize their technology transfers, likely share perceptions of trustworthiness similar to their

compatriots. Consistent with this assumption, we therefore expect that a Spanish subsidiary of

an Italian MNE and a Spanish subsidiary of a Dutch MNE have a different governance of their

foreign R&D inputs in a way that is consistent with national differences in perceived

trustworthiness. The reason is that there is ample evidence that trust in the foreign country

where the subsidiary is located increases the autonomy of the subsidiary with respect to the

business group (Aghion and Tirole, 1997; Bloom et al., 2012; Kastl et al., 2013; Cingano and

Pinotti, 2016). This in turns can reduce technology transfers within the business group and

induce more technology procurement via market channels. This argument is consistent with the

theory of transaction cost economics (Williamson, 1985, 1993), which predicts that a reduction

in mutual dependence favors market exchanges and reduces vertical integration.

An alternative possibility is that trust into the host economy might reduce the concerns

about a subsidiary’s ex post opportunistic behavior and thereby increase technology transfers

within the business group. The reason is that when a firm transfers sensitive technology to a

foreign subsidiary, this knowledge is not completely excludable. It can be leaked or misused by

the employees of the subsidiary. High trustworthiness suggests that such opportunistic behavior

is less likely than when trustworthiness is low. Moreover, if within-firm technology flows

substitute knowledge from the market, higher perceived trustworthiness might reduce the

imports of technologies that the subsidiary obtains from market channels. The main

4
contribution of this paper is to provide an empirical assessment of the relative importance of

these two possibilities.

We use a unique panel dataset on European firms operating in Spain. It provides us with

exhaustive information on firms’ R&D imports by provider. We combine these data with the

Eurobarometer survey’s measure of trust between citizens of European countries. Our dataset

includes information about 907 affiliates from European companies operating in Spain for the

period 2005-2012. We have two measures of international technology transfers: R&D imports

from companies that belong to the same business group (called from the group), and R&D

imports from international providers located abroad that do not belong to the same business

group (called from the market). These data provide, to our knowledge, the most detailed firm-

level panel data information on R&D transactions by suppliers worldwide.

European companies operating in Spain provide a good testing case for our research

question. The reason is that Spain has been one of the main receivers of foreign direct

investment from the European Union, and the risk of expropriation was very small in the period

analyzed (OECD, 2015). In addition, countries which belong to the EU share a common

institutional environment and intellectual property rights, which helps us to isolate the influence

of trust on technology transfers. Moreover, our nationally representative measure of trust in

Spaniards is exogenous to unobservable firm characteristics because trust is not measured in

the headquarters of the subsidiary by asking the managers directly and it precedes the years of

the firm technology transfers.1

We deploy two identification strategies in this paper. The first strategy uses differences

of trust that citizens from the countries of the parent companies have in Spaniards. In this way,

we explore the cross-sectional variation of the data. The second identification strategy includes

firm fixed effects in our regressions to analyze the effect of trust for subsidiaries that experience

1
In the Eurobarometer, the question on bilateral trust was not asked after 1997.

5
changes in the location of their parent company. This implies that we calculate results for

within-firm variation of trust in Spaniards. Our findings are robust after including time-variant

firm variables and extensive controls for heterogeneous country characteristics.

Our results show that perceived trustworthiness in Spaniards significantly affects

technology transfers. High levels of perceived trustworthiness in Spaniards raise international

technology transfers from the market and reduce imports from the business group. We find in

our cross-section estimations that increasing trustworthiness to the fourth quartile of the

trustworthiness distribution reduces international technology transfers from the business group

by 27.4%. The results from the estimations based on the within-firm trustworthiness variation

are very similar for increases in trustworthiness beyond the average level of trustworthiness

placed in Spaniards. This suggests that imports from the business group are important for low

levels of trustworthiness and they decline once a high level of trustworthiness is reached.

In cross-section estimations, we find that increasing trustworthiness in Spaniards raises

imports from the international market for the entire distribution of trustworthiness in Spaniards,

but this effect is concentrated in the fourth quartile of trustworthiness in our within-

trustworthiness estimations. In these estimations, we find that increasing trustworthiness to the

fourth quartile of the trustworthiness distribution increases technology transfers from the

market by 22.9%. Overall, our findings suggest that trustworthiness is important for

understanding the governance of technology transfers.

The paper is organized as follows. Section 2 presents the related literature and the

theoretical considerations. Section 3 describes data sources, the construction of the main

variables, and some empirical regularities. Section 4 shows the econometric specification and

the description of the control variables. The results are presented in Section 5. Section 6

concludes.

6
2. Related literature and theoretical considerations

One fundamental reason argued in the literature for the positive relationship between trust

and economic growth is that agents with higher levels of trust are likely to invest and trade more

than those with less trust (Fehr, 2009). Within organizations, higher levels of trust have been

associated with increasing workplace performance (Brown et al., 2015), innovation (Godart et

al., 2017) and firm growth (Bloom et al., 2012). A mechanism behind these relationships is that

principals who trust their agents more favor delegation and induce higher efforts from their

agents. For example, Cingano and Pinotti (2016) present evidence from a sample of Italian

firms that companies located in regions with higher levels of trust have larger value-added

shares in delegation-intensive industries relative to other industries. Kastl et al. (2013) find that

delegation increases R&D expenditures for a sample of Italian firms. Bloom et al. (2012) show

for a sample of European MNEs that higher levels of bilateral trust between a multinational’s

country of origin and a subsidiary’s country increases decentralization and firm size. An

important difference in our approach with respect to this literature is that we focus on how the

governance of technology imports depends on the host country’s perceived trustworthiness of

the subsidiary.

In synthesis, these findings suggest that higher levels of perceived trust increase

delegation for the provision of technology to the subsidiary. According to transaction cost

economics (Williamson, 1985, 1993a), one consequence of a reduction of the dependence of

the subsidiary on the group is a reduction of trade of technology within the firm and an increase

in procurement from the market. The reason is that vertical integration is more likely when

mutual dependence between sellers and buyers is high, and high trust reduces such dependency.

Therefore, trust might shape the decision of foreign subsidiaries to source their R&D imports

from the international market or from within the business group. By providing evidence for this

7
channel, our paper contributes to a better understanding of how knowledge is obtained and

organized through foreign direct investment.

An alternative argument that might compete with the previous explanations is that trust

can reduce concerns about the subsidiary’s ex post opportunistic behavior This logic might be

particularly important in the case of technology because knowledge can be leaked. For example,

Lai et al (2009) argue that R&D is not outsourced as much as manufacturing goods because

outsourcing R&D might lead to the potential leakage of trade secrets. Moreover, market failures

might likely arise in high R&D intensive goods. In this line, Siddharthan and Kurma (1991)

find for US multinationals that internalization is high in R&D and skill intensive industries

given to the product novelty and buyer’s uncertainty.

Our paper also advances the literature on the determinants of intra-firm trade by showing

that trust is key to discerning the source of foreign flows of technology within firms. The

literature has predominantly focused on the role of contractual incompleteness and the hold-up

problem for the choice between outsourcing and vertical integration. 2 The theoretical

arguments of Antràs (2003) and Antràs and Helpman (2008) are based on the property rights

theory of Grossman and Hart (1986). Their emphasis is on the role of asset ownership that

allows allocating residual rights of control in case there are contingencies in the trade

relationship. From an empirical perspective, Ulset (1996) analyzes determinants of internal and

external R&D projects and their relationship with contractual incompleteness. Nunn and Trefler

(2013) argue that intra-firm trade is positively related to non-contractible headquarter inputs

while Bernard et al. (2010) show that it is negatively related to product contractibility and

country governance quality.3 Moreover, Bernard et al. (2010) consider that country governance

might mitigate the effect of product contractibility and thus reduce intra-firm trade. In contrast

2
See Lafontaine and Slade (2007) for a review of the literature.
3
Country governance quality is a variable constructed from the World Bank governance indicators.

8
to this literature, we study how, for a given level of product contractibility, trustworthiness in

the host country induces intra-firm and market trade of technology.

Previous literature on intra-firm trade (Helleiner and Lavergne, 1979; Siddharthan and

Kumar, 1990) suggests that transaction costs and market failures are expected to increase

within-firm trade and decrease open market transactions. Intra-firm trade is substituted over

time by arm-length transactions when the market grows, and transaction costs decrease. In

contrast to our paper, these articles do not analyze the role of trustworthiness on intra-firm and

market trade.

Finally, our paper also sheds new light on technology transfers within MNEs. Several

articles consider transfers of intangibles and technology within MNEs essential to

understanding the existence of MNEs (Teece, 1977; Markusen, 1984; Teece, 1986; Atalay et

al., 2014; Ramondo et al., 2016). Branstetter et al. (2006) find that technology transfer payments

within MNEs increase after intellectual property right reforms. One possible reason is that the

likelihood of knowledge leakages decreases after a strengthening of intellectual property rights.

The novelty of our paper is that we provide evidence of one type of technology transfer within

MNEs, namely, R&D imports. Moreover, we show that R&D imports within the group are

quantitatively larger than technology obtained from the market. In addition, our results suggest

that transfers within MNEs are key in low-trust environments.

3. Data, main variables and some empirical regularities

The aim of this paper is to assess the effect of perceived trustworthiness in the host country

on technology transfers within the business group and from the market for subsidiaries from

foreign MNEs. Our source of firm-level data comes from a survey of firms operating in Spain

(Panel de Innovación Tecnológica, PITEC). It is a panel database constructed by the Spanish

9
National Institute of Statistics on the basis of annual responses to the Community Innovation

Survey (CIS) administered to a representative sample of Spanish firms.4 In the survey, each

company provides information on some of its economic data, such as sales or number of

employees, its ownership structure, the location of its parent company and, most importantly

for our research question, very detailed information on firms’ imports of technology

distinguished by provider.5

We conduct the empirical analysis for the years 2005 to 2012. We also use country-level

data from different sources, which we explain below, to control for characteristics of the parent

company’s country. Based on the availability of data on our trust measure, we use a panel of

907 firms that are subsidiaries of foreign European MNEs operating in Spain during the sample

period. 6 The panel contains an average of seven observations per firm, with the parent

companies of these subsidiaries located mostly in France (27.6%), Germany (24.5%), the

Netherlands (10.7%), the UK (10%), and Italy (7.7%), as shown in the first column of Table 1.

3.1. International technology transfers from inside and outside the business group

Our main interest is to analyze to what extent trustworthiness influences international

technology transfers within the business group and from the market. Our measures of

international technology transfers are R&D services acquired abroad by the subsidiary within

the business group and from the market. In the survey, each company indicates its R&D

acquisitions, that is, its purchases of R&D services.7 R&D acquisitions are defined in the survey

as:

4
The PITEC survey is specifically designed to analyze R&D and other innovating activities following the
recommendations of the OSLO Manual on performing innovation surveys (see OECD 2005). Details on PITEC
and data access guidelines can be obtained at http://icono.fecyt.es/PITEC/Paginas/descarga_bbdd.aspx.
5
The questions we quote below are from the English version of the CIS questionnaire. These questions are the
exact equivalent of those in the Spanish questionnaire.
6
In the Eurobarometer, there is information on trust in Spaniards from citizens from 17 European countries. In our
firm-level sample, there are 907 firms from these countries.
7
R&D services are defined in the survey as: “Creative work to increase the volume of knowledge and to create
new or improved products and processes (including the development of software)”.

10
 “Acquisitions of R&D services outside the firm through contracts, informal agreements,

etc… Funds to finance other companies, research associations, etc… that do not directly imply

purchases of R&D services are excluded”.

The data allow for the disaggregation of R&D acquisitions into domestic purchases (in

Spain) or imports (from abroad). With this information, we construct the variables R&D imports

within the group and R&D imports from the market. Both variables are the logarithm of imports

of R&D from companies that belong to the same business group and R&D imports from

providers that do not belong to the same business group, respectively, after dividing by the

number of employees. Measures similar to our measure of international technology transfers

within the group are used by Hu et al. (2005) and Branstetter et al. (2006), who take the

expenditures on disembodied technology purchased from foreign providers as a measure of

foreign technology transfers. In contrast to previous research, we can account for the exact

amount of technology imports of a given firm within and outside the business group, and

thereby compare technology transfers within a vertical integration relationship or through

market channels at the firm level.

3.2. Trust in the host country

Our main independent variable is trust in Spaniards by citizens from the country where

the parent company is located. Trust data are constructed with data from several waves of the

Eurobarometer survey, as in Guiso et al. (2009) and Bloom et al. (2012).8 In the Appendix of

this paper, we include full details on all country data sources. In the survey, citizens from

different countries (mostly from the European Economic Area) are asked:

8
The Eurobarometer is available at http://ec.europa.eu/COMMFrontOffice/publicopinion/index.cfm. We collect
information from the following waves of the Eurobarometer where this question is asked: 1986, 1990, 1993, 1994,
1995, 1996, and 1997. For Eastern European countries, we collect information from the Central and Eastern
European Barometer for the year 1990. Our measure is calculated as the average trust for these years.

11
 “I would like to ask you a question about how much trust you have in people from various

countries. For each, please tell me whether you have a lot of trust, some trust, not very much

trust or no trust at all”.

A number that varies from 1 = “no trust at all” to 4 = “a lot of trust” is assigned to each

answer. Our measure of trust in Spaniards is computed as the average trust that citizens of the

MNE’s headquarter country of a subsidiary have in Spaniards. We remove country fixed effects

by dividing this average by the mean trust of each country. The main advantage of the trust

variable as elicited in the Eurobarometer is that it reflects the perception of the average citizen

of a given country. Therefore, it also likely reflects the thoughts and attitudes of managers and

middlemen of a firm (Guiso et al., 2009). Moreover, since the Eurobarometer survey is not

directly conducted in the headquarters of the subsidiaries that we have in our sample and it

precedes the years of the firm technology transfers, our measure of trust is exogenous to

unobservable firm characteristics. Put differently, if managers of firms in our sample had been

asked about how much trust they have in their subsidiaries, omitted variables would be

correlated to this measure of trust and technology transfers. For example, if technology transfers

had been very profitable for the firm, managers would likely believe more in the trustworthiness

of the subsidiary and this would overstate the causal link between trust and technology transfers.

On the flipside, if technology transfers had been unprofitable, the causal link between trust and

technology transfers would be understated.

The key features of our measure of trust are that there is variation across countries in the

sample, as shown in the second column of Table 1, and there are some changes in the

headquarter location. Most countries trust Spaniards more than they trust other countries, on

average. This is particularly relevant for Greek, Austrian and Italian citizens. Citizens from

these countries trust Spanish citizens from 15.1% to 21.5% more than they trust citizens from

other countries in the sample. In contrast, citizens from Norway, Slovakia and the United

12
Kingdom trust Spain from 10.2% to 15% less than they trust people from other countries, on

average.

TABLE 1

3.3. Empirical regularities: Trust in Spaniards and international technology transfers

Figure 1 provides a sense of the relationship between trust and international technology

transfers within the group and through market channels. We divide R&D imports into four trust

quartiles. On average, imports within the group are around two and a half to three times larger

than imports from the international market. The distribution of imports within the group across

quartiles suggests that as trust increases from the first to the third quartile, imports within the

group increase. However, further increases in trust dramatically decrease R&D imports within

the group. By contrast, imports from the international market increase as trust increases, except

from the second to the third quartile, where there is a small decline. These features suggest that

there is an inverted-U relationship between trustworthiness and imports within the group, and

a positive relationship between trustworthiness and imports from the international market. We

now turn to use econometric techniques to analyze whether these relationships are robust after

controlling for covariates.

FIGURE 1

4. Econometric specification and control variables

Our main goal is to estimate the effect of trustworthiness in the subsidiary on technology

transfers within the business group and through market channels. Our baseline specifications

are as follows:

𝑅&𝐷 𝑖𝑚𝑝𝑜𝑟𝑡𝑠 𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦 𝑤𝑖𝑡ℎ𝑖𝑛 𝑓𝑖𝑟𝑚 (𝑓𝑟𝑜𝑚 𝑚𝑎𝑟𝑘𝑒𝑡)𝑖𝑗𝑡 = 𝛽1 + 𝛽2 𝑇𝑟𝑢𝑠𝑡𝑗 + 𝛽3 X𝑖𝑡−1 +

+𝛽4 Z𝑗𝑡−1 + 𝑑𝑡 + 𝜖𝑖𝑡 , (1)

13
where the sub-indices refer to company i from country j (where the parent company is located)

in year t. The variable 𝑇𝑟𝑢𝑠𝑡𝑗 denotes trustworthiness in Spaniards from citizens from the

country where the parent company is located; X𝑖𝑡−1 are a set of firm characteristics; Z𝑗𝑡−1 are

variables that control for the features of the country where the parent company is located and

economic and cultural similarities with respect to Spain, which we explain in detail below; 𝑑𝑡

are year fixed effects, and 𝜖𝑖𝑡 is the error term. The coefficient of interest is 𝛽2, which captures

the effect of trustworthiness on a subsidiary’s R&D imports.

We include the following firm characteristics as control variables: labor productivity,

the average wage in the R&D department and an indicator that takes the value 1 if the company

has applied for patents to control for absorptive capacity and skill mix, and 0 otherwise. We

also include market power to control for competition and capital per employee to control for

physical investments. Moreover, we include two variables to measure the degree of innovative

or adopted R&D within the subsidiary given that innovative R&D and outsourcing of

technology might be complementary (Odagiri, 1983). The first variable that we add is

innovative R&D intensity, which accounts for the percentage of sales of products new to the

market. The second variable is an indicator that takes the value 1 if the firms have not

undertaken any product or process innovation, and 0 otherwise. Finally, we add export status

and the degree of internationalization of the sector where the subsidiary operates because trade

can induce companies to engage in other globalization strategies (e.g., Tomiura, 2007). The

sectoral openness is calculated as sectoral imports plus exports over sectoral production (these

sectoral data come from the OECD).

We add the following variables as country controls: Companies that come from high-tech

countries can have access to a wider technological network and profit from updated

technologies, which can enhance technology transfers. Following these arguments, first, we

include the variable R&D expenditures as percent of GDP in our specification. The data come

14
from the World Bank. Second, we include the logarithm of the GDP per capita (data from the

IMF). Firms’ R&D imports might also depend on tax differentials (Devereux and Griffith,

1998). For this reason, we include a measure of corporate taxes as the ratio of corporate income

taxes of a given country over corporate income taxes of Spain. The data come from the tax

database of the OECD. If corporate taxes are higher in Spain than in the partner country, MNEs

might increase imports within the group to reduce taxable profits in the high-tax country.

Therefore, we might expect a positive relationship between relative corporate taxes and R&D

imports within the group. Moreover, we control for R&D policy incentives (R&D taxes and

other R&D related subsidies). This variable is constructed as the percentage of business and

enterprise R&D financed by the government of a given country minus the percentage of

business and enterprise R&D financed by the government of Spain. These data come from the

OECD. Finally, we add the physical distance between capitals. Keller and Yeaple (2013) argue

that complex technologies become costly to transfer as transport costs increase, which would

decrease both types of technological transactions. We use the distance between capitals as a

proxy for transport costs. 9 In the Appendix of the paper, we document detailed variable

definitions, data sources, and summary statistics (Table A1). Table OA1 in the On-line

Appendix documents the correlation between the variables. It turns out that most variables are

only weakly correlated.

We employ two identification strategies. The first strategy uses the differences of trust

that the citizens from the countries of the parent companies have in Spaniards. In this way, we

explore the cross-sectional variation of the data. The second identification strategy includes

firm fixed effects in our regressions. In this way, we can analyze the effect of trust when there

are changes in the location of the parent company. There are 161 foreign subsidiaries whose

9
This is the shortest distance between the two capitals on the surface of the earth, which is measured as a sphere.
This variable is in logarithms. The data come from CEPII.

15
headquarters changed their location during the sample period. Hence, our trust coefficient is

identified from within-firm variation of trust while controlling for all time-invariant firm

characteristics that might affect R&D imports.

5. The effect of trust on technology transfers:

5.1. Our main result: Distinguishing between the main channels

In this section, we analyze the relationship between trustworthiness and international

technology transfers, distinguishing between within the firm or through international market

channels. We present the results in Table 2. In panel A, the dependent variable is the natural

logarithm of R&D imports from the business group and in panel B, the dependent variable is

the natural logarithm of imports from the market.

In columns 1, 3, 5 and 7, we show results for our measure of trustworthiness in Spaniards.

In columns 2, 4, 6 and 8, we include indicator variables for each quartile of trustworthiness in

Spaniards, to account for non-linear effects. We report estimates from four specifications: (i) in

columns 1 and 2, we do not include any firm or country controls; (ii) in columns 3 and 4, we

include country controls; (iii) in columns 5 and 6, we add firm controls; and (iv) in columns 7

and 8, we include firm fixed effects to measure the effect of trustworthiness for firms whose

parent company changes its location over the sample period. In all regressions in both panels,

we include industry and year fixed effects. All standard errors are clustered at the country level.

In column 1a, we observe that trust is negatively related to R&D imports from the

business group although this effect is not precisely estimated. Once we distinguish between

trust quartiles, we find a negative and highly significant effect of trust in Spaniards on imports

from the business group for MNEs at the highest level of trust in Spaniards. This relationship

is robust after the inclusion of country or firm controls. The magnitude of the effect is large: in

the most conservative estimations in column 6a, MNEs within the fourth quartile of trust

16
decrease technology transfers from the market by 27.4% as compared with MNEs in the first

quartile of trust in Spaniards. This implies that subsidiaries from countries with the highest trust

in Spaniards have fewer technology transfers within the firm than subsidiaries from other

countries. In columns 7a and 8a, we include firm fixed effects to measure how within-firm trust

variation affects R&D imports. The coefficients from columns 7a are negative and statistically

significantly different from zero. Focusing on the quartile measures of trust, in columns 8a, we

note that the negative effect is concentrated in MNEs from countries with above average trust

in Spaniards (3rd and 4th trust quartile).

TABLE 2

With respect to the relationship between trustworthiness in Spaniards and R&D imports

from the market in panel B, we show in columns 1b to 6b that trust is always strongly positively

related to R&D imports from the international market. For example, the coefficient of trust in

column 5b indicates that an increase in trust by one unit increases imports from the international

market by 88.3%. Once we include firm fixed effects in columns 7a and 8a, we find that the

positive relationship is only significant for the highest quartile of trust.

With respect to the country controls, the estimated coefficient of the GDP per capita

variable is negative and significant in panel A and positive and significant in panel B. R&D

expenditures as percent of GDP is negative and significantly related to imports from the

international market in panel B. The effect is the opposite in the case of tax similarities, which

is positive and significant at conventional levels in panel B. Among the firm-level controls,

patents, average physical investment, and being an exporter are significant and positively

related to R&D imports from the group. Average salary in R&D, patents, innovative R&D

17
intensity, labor productivity and average physical investment are significant and positively

related to imports from the international market.10

Overall, the results in panels A and B show a pattern of opposite-signed effects of trust

on technology transfers from the group and from the market, which suggests that as trust

increases, there is a reduction in the knowledge flows within the firm and an increase in the

technology imported from external providers.

In Table 3, we present results from alternative econometric specifications. In order to

account for observations with zero R&D imports, we present estimates using a random effect

Tobit model in columns 1 and 2 and a fixed effect Poisson model in columns 3 and 4. In all

specifications, we include industry and year fixed effects and country and firm controls in all

regressions. The key messages remain unchanged.

TABLE 3

Moreover, to account for potential omitted variables that might have overstated the

estimated trust coefficients, in Table OA2 in the On-line Appendix, we include additional

country controls and, in Table OA3 in the On-line Appendix, we drop countries considered tax

havens according to Spanish jurisdiction.11 We include three institutional indicators that capture

different dimensions of rule violations and law enforcement in the country of the parent

company (the source of these variables is the World Bank): the rule of law; government

accountability and control of corruption. We add these indicators because the trust in Spaniards

variable might be reflecting perceived differences in rule violations (Bloom et al., 2012).

10
Although our results appear to be weak in terms of R2, F-tests reject in all cases the null hypothesis that all
independent variables in the model are not jointly significant in affecting the dependent variable.
11
This information is provided by the European Union at http://ec.europa.eu/taxation_customs/business/company-
tax/tax-good-governance/tax-good-governance-world-seen-eu-countries_en. We drop observations from Ireland,
whose average corporate taxes for the sample period was 12.5%, and Luxembourg because of the special low taxes
on dividends and capital gains.

18
 In all regressions, the results are consistent with our previous estimations in Table 2 and

support the positive relationship between trustworthiness and R&D imports from the

international market and the negative effect of trustworthiness on imports from the business

group. This suggests that our results are not biased by omitted variables related to country

institutional differences or tax havens.12

5.2. The joint decision of importing from the business group and from the market

To shed light on potential differences for firms that simultaneously import R&D from the

business group and from the market, in Table 4 we consider the percentage of R&D imports

from the business group over total R&D imports as the dependent variable. Our estimations are

for the sample of firms with positive total R&D imports. In the following tables, we follow the

same specification as in Table 2.

TABLE 4

Our results in Table 4 show that the coefficient of trust is negative and statistically

significant in columns 1 and 5. However, in the specifications in columns 3 and 7, trust is not

significantly different from zero. Once we differentiate across trust quartiles, we observe a

strong negative relationship for the second and fourth quartile of trust, which suggests that as

trust increases, there is a decrease in the percentage of imports from the group, but this effect

is not linear. This result remains once we include country and firm controls in columns 4 and 6.

In columns 7 and 8, once we add firm fixed effects, the coefficient of trust remains negative

and statistically significant for the second quartile of trust. These results support the conclusion

12
Note that, once we exclude countries considered as tax havens, corporate taxes are relatively similar across
countries in our sample for the analyzed period.

19
of a negative effect of trust on the percentage of imports from the business group and the

consequently positive effect for imports from the international market.

5.3. Distinguishing between international market channels

In order to gain further insights on the effects of trust on R&D transactions through the

international market, we explore the heterogeneous effect of trust on different stratifications of

international market channels. In Table OA4 in the On-line Appendix, we report the

classification and definitions of the different subcategories of R&D imports. In Table 5, we

present estimates based on three different international market channels. Panel A reports

evidence for foreign private firms. Panel B reports results for foreign universities and Panel C

presents evidence for other foreign sources not previously included.13

TABLE 5

In all cases the results confirm a positive relationship between trust and imports of R&D

from the different foreign channels. In Panel A, the estimated coefficients in the different

specifications are quite similar to those in Table 2 for total R&D imports from the foreign

market. This suggests that trust is key in obtaining new knowledge from other foreign private

companies that are not part of the business group. In Panels B and C, estimated coefficients are

significantly lower than in Panel A. This implies that trust in the subsidiary is less important

when the subsidiary is acquiring R&D services from foreign universities or other foreign

sources than when they acquiring them from other foreign private firms. One possible reason

is that private firms generate more commercial research than universities. For example, Spencer

(2001) finds that corporate research is more influential than university research for Japanese

13
For the average firm, R&D imports from foreign private firms account for most R&D imports from the foreign
market (98.1% on average). Other foreign sources include foreign public administrations, foreign non-profitable
organizations and non-specified foreign sources.

20
firms’ commercial applications. Therefore, a higher level of delegation might be needed for the

subsidiary to coordinate with foreign private firms than with foreign universities.

7. Summary and concluding remarks

In this paper, we studied the effects of trust on technology transfers. This is important

for understanding how knowledge is organized within MNEs. We estimated determinants of

R&D imports within the group and through market channels for affiliates from European firms

operating in Spain. We find evidence that trustworthiness in Spaniards is negatively related to

technology transfers within the business group and positively related to acquisitions through

market channels. These results are consistent with transaction cost economics (Williamson,

1985, 1993) and support the hypothesis of the importance of vertical transactions in low-trust

environments, and on the role of trust to enhance decentralization and delegation to the affiliate.

Our findings suggest that an increase in trustworthiness may significantly change the channels

through which knowledge flows, and as a consequence, influence innovation and reduce

technological dependence on the foreign MNE.

21
References

Aghion, P., Tirole, J., 1997. Formal and real authority in organizations. The Journal of Political
Economy 105(1), 1-29.

Antràs, P., 2003. Firms, contracts, and trade structure. Quarterly Journal of Economics 118(4),
1375-1418.

Antràs, P., Helpman, E., 2008. Contractual frictions and global sourcing. In: Helpman, E.,
Verdier, T., Marin, D. (Eds.). The Organization of Firms in a Global Economy.
Cambridge, MA: Harvard University Press, 9-54.

Atalay, E., Hortaçsu A., Syverson, C., 2014. Vertical integration and input flows. American
Economic Review 104(4), 1120-1148(29).

Bernard, A., Jensen, J.B., Redding, S., Schott, P.K., 2010. Intra-firm trade and product
contractibility. American Economic Review 100(2), 444-448.

Bertrand, O., 2009. Effects of foreign acquisitions on R&D activity: Evidence from firm-level
data for France. Research Policy 38(6), 1021-1031.

Bloom, N., Sadun, R., Van Reenen, J., 2012. The organization of firms across countries. The
Quarterly Journal of Economics 127(4), 1663-1705.

Branstetter, L., Fisman, R., Foley, C.F., 2006. Do stronger intellectual property rights increase
international technology transfer? Empirical evidence from U. S. firm-level panel data.
The Quarterly Journal of Economics 121, 321-349.

Brown, S., Gray, D., McHardy, J., Taylor, K., 2015. Employee trust and workplace
performance. Journal of Economic Behavior and Organization 116, 361–378.

Cingano, F., Pinotti, P., 2016. Trust, firm organization, and the pattern of comparative
advantage. Journal of International Economics 100, 1–13.

Criscuolo, C., Haskel, J., Slaughter, M., 2010. Global engagement and the innovation activities
of firms. International Journal of Industrial Organization 28, 191–202.

Dearmon, J., Grier, K.K., 2009. Trust and development. Journal of Economic Behavior and
Organization 71(2), 210–220.

Devereux, M.P., Griffith, R., 1998. Taxes and the location of production: evidence from a panel
of US multinationals. Journal of Public Economics 68, 335–367.

22
Eaton, J., Kortum, S., 1996. Trade in ideas: Patenting and productivity in the OECD. Journal
of International Economics 40(3-4), 251-278.

European Commission, several years. Eurobarometer, available at

http://ec.europa.eu/public_opinion/index_en.htm

Fehr, E., 2009. On the economics and biology of trust. Journal of the European Economic
Association 7(2-3), 235-266.

Godart, O., Görg, H., Hanley, A., 2017. Trust-based work-time and product improvements:
Evidence from firm level data. Industrial and Labor Relations Review 70(4), 894-918.

Grossman, S.J., Hart, O.D., 1986. The costs and benefits of ownership: A theory of vertical and
lateral integration. Journal of Political Economy 94(4), 691-719.

Guiso, L., Sapienza, P., Zingales, L., 2008. Trusting the stock market. The Journal of Finance
63(6), 2557–2600.

Guiso, L., Sapienza, P., Zingales, L., 2009. Cultural biases in economic exchange? The
Quarterly Journal of Economics 124(3), 1095-1131.

Helleiner, G., Lavergne, R., 1979. Intra-firm trade and industrial exports to the United States.
Oxford Bulletin of Economics and Statistics 41(4), 297-311.

Hu, A., Jefferson, G., Jinchang, Q., 2005. R&D and technology transfer: Firm-level evidence
from Chinese industry. The Review of Economics and Statistics 87(4), 780-786.

Kastl, J., Martimort, D., Piccolo, S., 2013. Delegation, ownership concentration and R&D
spending: evidence from Italy. Journal of International Economics 61(1), 84–107.

Keller, W., 2004. International technology diffusion. Journal of Economic Literature 42(3),
752-782.

Keller, W., Yeaple, S.R., 2013. The gravity of knowledge. The American Economic Review
103(4), 1414-1444.

Knack, S., Keefer, P., 1997. Does social capital have an economic payoff? A cross-country
investigation. Quarterly Journal of Economics 112(4), 1251–1288.

Lai, E., Riezman, R., Wang, P., 2009. Outsourcing of innovation. Economic Theory 38, 485-
515.

23
Lafontaine, F., Slade, M., 2007. Vertical integration and firm boundaries: The evidence. Journal
of Economic Literature 45(3), 629-685.

La Porta, R., López-de-Silanes, F., Shleifer, A., Vishny, R., 1999. The quality of Government.
Journal of Law, Economics and Organization, 15(1), 222-279.

Markusen, J., 1984. Multinationals, multi-plant economies, and the gains from trade. Journal
of International Economics 16(3–4), 205–26.

Nunn, N., Trefler, D., 2013. Incomplete contracts and the boundaries of the multinational firm.
Journal of Economic Behavior and Organization 94(1), 330-344.

Odagiri, H., 1983. R&D expenditures, royalty payments, and sales growth in Japanese
manufacturing corporations. Journal of Industrial Economics 32(1), 61–71.

OECD, 2005. Oslo Manual: Guidelines for Collecting and Interpreting Innovation Data, 3rd
Edition. Paris: OECD Publishing.

OECD, 2015. Country risk classification, available at http://www.oecd.org/tad/xcred/crc.htm

Panel de Innovación Tecnológica (PITEC). Information available at http://sise.fecyt.es/sise-

public-web/mostrarCarpetaEstudiosInformes.do

Ramondo, N., Rappoport, V., Ruhl, K., 2016. Intrafirm trade and vertical fragmentation in U.S.
multinational corporations. Journal of International Economics 98(1), 51-59.

Siddharthan, N., Kumar, N., 1990. The determinants of inter-industry variations in the
proportion of intra-firm trade: The behaviour of U.S. multinationals.
Weltwirtschaftliches Archiv 126(3), 581-591.

Spencer, J., 2001. How relevant is university-based scientific research to private high-
technology firms? A United States–Japan comparison. Academy of Management
Journal 44(2), 432-440.

Stolle, D., 2002. Trusting strangers - the concept of generalized trust in perspective.
Österreichische Zeitschrift für Politikwissenschaft 31(4), 397-412.

Teece, D., 1977. Technology transfer by multinational firms: the resource cost of transferring
technological know-how. The Economic Journal 87(346), 242-261.

Teece, D., 1986. Transactions cost economics and the multinational enterprise: An assessment.
Journal of Economic Behavior and Organization 7(1), 21-45.

24
Tomiura, E., 2007. Foreign outsourcing, exporting, and FDI: A productivity comparison at the
firm level. Journal of International Economics 72(1), 113–127.

Ulset, S., 1996. R&D outsourcing and contractual governance: An empirical study of
commercial R&D projects. Journal of Economic Behaviour and Organization 30(1), 68-
82.

Williamson, O., 1971. The vertical integration of production: Market failure considerations.
The American Economic Review 61(2), 112-123

Williamson, O., 1981. The Modern Corporation: Origins, Evolution, Attributes. Journal of
Economic Literature 19(4), 1537-1568.

Williamson, O., 1985. The Economic Institutions of Capitalism. New York, NY: The Free
Press.

Williamson, O., 1993a. Opportunism and its critics. Managerial and Decision Economics 14(2),
97-107

Williamson, O., 1993b. Calculativeness, trust, and economic organization. The Journal of Law
and Economics 36(1), 453-486.

Zak, P. J., Knack., S., 2001. Trust and growth. The Economic Journal 111(470), 295–321.

25
FIGURE AND TABLES

Figure 1: Trust in Spaniards and R&D imports by trust quartiles

.8
.6

From the group

.4

From the Market

.2
0

Q1 Q2 Q3 Q4
Quartiles of trust in Spaniards

26
Table 1: Distribution of subsidiaries by country of origin and trust in Spaniards.
Country Observations Trust
Austria 56 1.172
Belgium 186 1.055
Czech Republic 1 0.913
Denmark 112 0.987
Finland 77 1.062
France 1,598 1.094
Germany 1,418 0.966
Greece 1 1.215
Ireland 46 1.028
Italy 447 1.151
Luxembourg 229 1.150
Netherlands 619 1.012
Norway 69 0.850
Portugal 142 1.112
Slovakia 1 0.894
Sweden 202 1.084
UK 578 0.898
Total 5782
Notes: The data include observations from all PITEC dataset firms that are subsidiaries of foreign MNEs during the
period 2005-2012. Trust is the average trust that citizens of the MNE’s headquarter country of a subsidiary have in
Spaniards, dividing by the mean trust of each country.

27
Table 2: Trust and R&D imports distinguishing between imports from the business group and the market
Panel A: Dependent variable R&D imports from business group
(1a) (2a) (3a) (4a) (5a) (6a) (7a) (8a)
Trust -0.552 -0.375 -0.326 -1.333**
(0.494) (0.453) (0.420) (0.582)
2nd quartile of trust -0.071 -0.090 -0.061 -0.096
(0.109) (0.110) (0.128) (0.111)
3rd quartile of trust 0.075 0.094 0.132 -0.403*
(0.055) (0.072) (0.091) (0.199)
4th quartile of trust -0.314*** -0.330*** -0.274** -0.194*
(0.057) (0.105) (0.099) (0.094)
Observations 5,782 5,782 4,814 4,814 4,804 4,804 4,804 4,804
R-squared 0.075 0.078 0.078 0.080 0.084 0.085 0.012 0.013
Panel B: Dependent variable R&D imports from the international market
(1b) (2b) (3b) (4b) (5b) (6b) (7b) (8b)
Trust 1.009*** 0.771*** 0.883*** 0.463
(0.276) (0.214) (0.140) (0.563)
2nd quartile of trust 0.271*** 0.275*** 0.255** 0.200
(0.075) (0.082) (0.091) (0.122)
3rd quartile of trust 0.125*** 0.088*** 0.120** -0.077
(0.022) (0.027) (0.046) (0.119)
4th quartile of trust 0.291*** 0.246*** 0.229*** 0.216*
(0.036) (0.046) (0.051) (0.118)
Observations 5,782 5,782 4,814 4,814 4,804 4,804 4,804 4,804
R-squared 0.046 0.050 0.088 0.091 0.106 0.107 0.015 0.016
Country controls yes yes yes yes yes yes
Firm controls yes yes yes yes
Firm FEs yes yes
Industry FEs and year FEs in all regressions
Notes: The dependent variable in panel A is R&D imports from the business group. The dependent variable in panel B is R&D imports from the international market. Country
controls are GDP per capita, geographical distance, R&D as percentage of GDP, R&D policy incentives and corporate taxes. Firm controls are average salary in R&D, patents,
labor productivity, market power, exporter, innovative R&D intensity, innovative indicator, average physical investment and industry internationalization. For exact definitions
and sources of all variables, see the Appendix. Estimated standard errors clustered at the country level are in parentheses. All estimations are OLS. * p < 10%, ** p < 5%, ***
p < 1%.

28
Table 3: Robustness checks. Tobit and Poisson models
Panel A: Dependent variable R&D imports from business group
(1a) (2a) (3a) (4a)
Estimation method: Tobit Tobit Poisson Poisson
Trust -1.183 -4.214***
(0.987) (1.115)
2nd quartile of trust -0.195 -0.422*
(0.227) (0.247)
3rd quartile of trust -0.050 -1.430***
(0.193) (0.356)
4th quartile of trust -0.435* -0.578**
(0.250) (0.243)
Observations 4,804 4,813 908 908
Panel B: Dependent variable R&D imports from the international market
(1b) (2b) (3b) (4b)
Estimation method: Tobit Tobit Poisson Poisson
Trust 1.179 1.478
(1.098) (1.475)
2nd quartile of trust 0.485* 0.767*
(0.257) (0.396)
3rd quartile of trust 0.164 -0.122
(0.171) (0.407)
4th quartile of trust 0.484* 1.098**
(0.281) (0.458)
Observations 4,804 4,813 581 581
Industry FEs, year FEs, country and firm controls in all regressions
Firm FEs Yes Yes
Notes: All independent variables are lagged one period. Country controls are GDP per capita, geographical
distance, R&D as percentage of GDP, R&D policy incentives and corporate taxes. Firm controls are average salary
in R&D, patents, labor productivity, market power, exporter, innovative R&D intensity, innovative indicator,
average physical investment and industry internationalization. For exact definitions and sources of all variables,
see the Appendix. Estimated standard errors clustered at the country level are in parentheses. All estimations are OLS.
* p < 10%, ** p < 5%, *** p < 1%.

29
Table 4: Exploring the effect of trust on the ratio of R&D imports from the group over total R&D imports
(1) (2) (3) (4) (5) (6) (7) (8)
Trust -0.747* -0.302 -0.658** 0.363
(0.373) (0.317) (0.259) (0.589)
2nd quartile of trust -0.172*** -0.195*** -0.245*** -0.227***
(0.046) (0.037) (0.045) (0.077)
3rd quartile of trust -0.023 0.000 -0.053 -0.156
(0.021) (0.019) (0.043) (0.092)
4th quartile of trust -0.289*** -0.252*** -0.299*** -0.079
(0.036) (0.053) (0.052) (0.067)

Observations 682 682 570 570 567 567 567 567

R-squared 0.212 0.248 0.330 0.351 0.381 0.397 0.121 0.128
Country controls yes yes yes yes yes yes
Firm controls yes yes yes yes
Firm FEs yes yes
Industry FEs and year FEs in all regressions
Notes: The dependent variable is the ratio between imports of R&D from the group over total R&D imports. All independent variables are lagged one period. Country controls
are GDP per capita, geographical distance, R&D as percentage of GDP, R&D policy incentives and corporate taxes. Firm controls are average salary in R&D, patents, labor
productivity, market power, exporter, innovative R&D intensity, innovative indicator, average physical investment and industry internationalization. For exact definitions and
sources of all variables, see the Appendix. Estimated standard errors clustered at the country level are in parentheses. All estimations are OLS.
* p < 10%, ** p < 5%, *** p < 1%.

30
Table 5: Trust and R&D imports from international market distinguishing between import channels
Panel A: Dependent variable R&D imports from foreign private firms
(1a) (2a) (3a) (4a) (5a) (6a) (7a) (8a)
Trust 0.894*** 0.668*** 0.759*** 0.493
(0.224) (0.188) (0.132) (0.566)
2nd quartile of trust 0.246*** 0.243*** 0.224** 0.120
(0.075) (0.076) (0.083) (0.099)
3rd quartile of trust 0.115*** 0.083*** 0.109** -0.107
(0.026) (0.027) (0.040) (0.122)
4th quartile of trust 0.253*** 0.200*** 0.186*** 0.198
(0.031) (0.039) (0.040) (0.118)
Panel B: Dependent variable R&D imports from foreign universities
(1b) (2b) (3b) (4b) (5b) (6b) (7b) (8b)
Trust 0.158*** 0.139*** 0.145*** 0.077
(0.036) (0.040) (0.045) (0.246)
2nd quartile of trust 0.039*** 0.051** 0.056* 0.117*
(0.009) (0.023) (0.031) (0.063)
3rd quartile of trust 0.026*** 0.019*** 0.010 -0.017
(0.005) (0.005) (0.019) (0.042)
4th quartile of trust 0.037*** 0.044** 0.054* 0.096*
(0.009) (0.015) (0.027) (0.055)
Panel C: Dependent variable R&D imports from other foreign sources
(1c) (2c) (3c) (4c) (5c) (6c) (7c) (8c)
Trust 0.094 0.083 0.103** -0.058
(0.075) (0.056) (0.040) (0.056)
2nd quartile of trust 0.034* 0.037* 0.033 0.020***
(0.017) (0.021) (0.023) (0.006)
3rd quartile of trust 0.007 0.004 0.013 -0.027
(0.009) (0.008) (0.011) (0.033)
4th quartile of trust 0.040*** 0.040*** 0.033** 0.008
(0.010) (0.014) (0.014) (0.007)
Country controls yes yes yes yes yes yes
Firm controls yes yes yes yes
Firm FEs yes yes
Industry FEs and year FEs in all regressions
Notes: For exact definitions and sources of all variables, see the Appendix. Estimated standard errors clustered at the country level are in parentheses. All estimations are OLS. * p < 10%, ** p <
5%, *** p < 1%.

31
APPENDIX

In this appendix, we define country variables, describe data sources that we use in our analysis
and present the descriptive statistics.

Definition of the variables

Trust: Our main independent variable is trust in Spaniards. It is the average trust that citizens
of the MNE’s headquarter country of a subsidiary have in Spaniards, dividing by the mean trust
of each country. The Eurobarometer survey provides information about this question. In the
survey, citizens from different countries (mostly from the European Economic Area) are asked:
“I would like to ask you a question about how much trust you have in people from various
countries. For each, please tell me whether you have a lot of trust, some trust, not very much
trust or no trust at all”. A number is assigned for each answer. We collect information from
the following waves of the Eurobarometer where this question is asked: 1986, 1990, 1993, 1994,
1995, 1996, and 1997 and construct the average of these values. For Eastern European
countries, we collect information from the Central and Eastern European Barometer for the year
1990.

GDP per capita is the natural logarithm of the GDP over the population of the MNE’s
headquarter country. The data are from the IMF (from
http://www.imf.org/external/pubs/ft/weo/2014/01/weodata/index.aspx) for the years 2005 to
2012. The variable “GDP per capita” ranges from 9.69 to 11.63 in our sample.

Geographical distance is the distance between the capital of the MNE’s headquarter country
and Madrid. The measure is in logarithms. The data are from CEPII
(http://www.cepii.fr/CEPII/en/bdd_modele/presentation.asp?id=6). For details on this
database, see Mayer, T. & Zignago, S. (2011), Notes on CEPII’s distances measures: the
GeoDist Database. CEPII Working Paper 2011-25.

R&D as % of GDP is the research and development expenditure as a percentage of the GDP
of the MNE’s headquarter country. The source of these data is the World Bank
(http://data.worldbank.org/indicator/GB.XPD.RSDV.GD.ZS?view=chart) for the years 2005 to
2012. The variable “R&D as % of GDP” is in logarithms and it ranges from -0.74 to 1.32.

Corporate taxes are the corporate income taxes of the MNE’s headquarter country over
Spanish corporate taxes. The data come from the OECD database
(http://stats.oecd.org//Index.aspx?QueryId=58204#), completed with data from KPMG Global
corporate tax rate tables (https://home.kpmg.com/xx/en/home/services/tax/tax-tools-and-
resources/tax-rates-online/corporate-tax-rates-table.html) for the years 2005 to 2012. The
variable ranges from 0.35 to 1.20.

R&D policy incentives is a variable that measures the difference between the percentage of
business and enterprise R&D that is financed by the government in the MNE’s headquarter
country and the percentage of business and enterprise R&D that is financed by the Spanish
government. It includes R&D tax credits and other subsidies related to business R&D. The data
come from the OECD (Main Science and Technology Indicators:
http://stats.oecd.org/OECDStat_Metadata/ShowMetadata.ashx?Dataset=MSTI_PUB&ShowO
nWeb=true&Lang=en) for the years 2005 to 2012. The variable ranges from -16.32 to -.049.

32
The following variables refer to characteristics of the MNE’s headquarter country used in the
robustness checks of the On-line Appendix. For the following three variables, the data come
from the Worldwide Governance Indicators (WGI) of the World Bank:
http://info.worldbank.org/governance/wgi/index.aspx#reports

Rule of law reflects perceptions of the extent to which agents have confidence in the rules of
society, and in particular the quality of contract enforcement and property rights. The index can
take values between -2.5 and 2.5. In our sample, it ranges from 0.35 to 1.99. We take data for
the years 2005 to 2012.

Accountability reflects perceptions of the extent to which a country's citizens are able to
participate in selecting their government, as well as freedom of expression, freedom of
association, and a free media. The index can take values between -2.5 and 2.5. In our sample,
it ranges from 0.80 to 1.76. We take data for the years 2005 to 2012.

Control of corruption reflects perceptions of the extent to which public power is exercised for
private gain, including both petty and grand forms of corruption, as well as "capture" of the
state by elites and private interests. The index can take values between -2.5 and 2.5. In our
sample, it ranges from -0.18 to 2.5. We take data for the years 2005 to 2012.

33
Table A1: Descriptive statistics

Variable Mean SD Observations

R&D imports from the group 0.538 1.933 8668
R&D imports from the market 0.267 1.305 8668
from foreign private firms 0.248 1.261 8668
from foreign universities 0.027 0.386 8668
from other foreign sources 0.019 0.322 8668
Trust 1.035 0.081 5782
Country controls
GDP per capita 10.674 0.311 8000
Geographical distance 7.552 0.719 8016
R&D as % of GDP 0.760 0.351 7716
Corporate taxes 1.032 0.184 8022
R&D policy incentives -8.857 3.726 7023
Firm controls
Average salary in R&D 5.033 5.177 8669
Patents 0.101 0.301 8669
Labor productivity 12.316 1.016 8668
Market power 1.000 2.374 8669
Exporter 0.723 0.447 8668
Average physical investment 6.508 3.664 8668
Innovative R&D intensity 8.082 19.919 7689
Innovative indicator 0.364 0.481 8638
Industry internationalization 0.134 0.116 9872
Institutional country indicators
Rule of law 1.520 0.380 8038
Accountability 1.294 0.269 8038
Control of corruption 1.551 0.498 8034
Notes: The data include observations from all PITEC dataset firms that are subsidiaries of foreign MNEs during the
period 2005-2012. R&D imports from the business group (market) is the natural logarithm of R&D imports from the
business group (market); the R&D imports from the international market can be disaggregated into those from foreign
private companies, from foreign universities and from other foreign sources. Trust is the trust that citizens of the MNE’s
headquarter country have in Spaniards. GDP per capita is the natural logarithm of the GDP over the population of the
MNE’s headquarter country. Geographical distance is the distance between the capital of the MNE’s headquarter
country and Madrid (in natural logarithm). R&D as % of GDP is the research and development expenditure as percentage
of the GDP of the MNE’s headquarter country. R&D policy incentives are the public incentives to business and enterprise
R&D of a given country minus those of Spain. Corporate taxes are the corporate income taxes of the MNE’s headquarter
country over Spanish corporate taxes. Average salary in R&D is the total salary in R&D over the number of employees
working in R&D of a firm. Patents is a dummy variable that takes the value one if the firm has applied for patents in the
current or previous two years. Labor productivity is the natural logarithm of the sales over the number of employees of
a firm. Market power is a firm’s sales relative to its industry. Exporter is an indicator that takes the value one if the firm
is an exporter. Average physical investment is the natural logarithm of a firm’s average physical investment over its
number of employees. Innovative R&D intensity is the percentage over sales of new products new to the market.
Innovative indicator is a dummy variable that takes the value one if the firm does not have any product or process
innovation. Industry internationalization is the degree of openness (imports plus exports over production) of the sector
where the firm operates. The following variables refer to characteristics of the MNE’s headquarter country used in the
robustness checks of the On-line Appendix: Rule of law reflects perceptions of the extent to which agents have
confidence in the rules of society, and in particular the quality of contract enforcement and property rights;
Accountability reflects perceptions of the extent to which a country's citizens are able to participate in selecting their
government; Control of corruption reflects perceptions of the extent to which public power is exercised for private gain.
For sources of all country variables, see the main text or the Appendix of the paper.

34